Human Trafficking Solution by Deep Learning with Keras and OpenCV

The entire world is currently facing problems related to higher rates of human trafficking today, and the research shows that the higher percentage of victims are among children. Although countless children are reported missing every year, many of them remain untraced. This paper presents a novel use of deep learning methodology for fighting against human trafficking and identifying the reported missing child from their photos, with the help of face recognition. The first step is to register the children in the common portal to be verified before crossing the border of their country of origin. Secondly, the public can upload photographs of suspected missing children into a common portal with landmarks and remarks. Uploaded photos are automatically compared with the registered photos of the missing children from the repository, and the classification of the input children’s images is performed. The photo with the best match will be selected from the database. To do that, a deep learning model is trained to correctly identify the missing child from the missing children image database provided, using the facial images uploaded by the public. The Convolutional Neural Network (CNN), a highly effective deep learning technique for image-based methodology, is adopted here for face recognition purposes. Face descriptors are extracted at 99.98% from the images using a pre-trained CNN model VGG16-Face deep architecture.